FULL PAPER 1 2 3 An efficient method for the surface functionalization of 4 5 luminescent quantum dots with lipoic acid-based ligands 6 7 Marcello La Rosa,[a,b,†] Tommaso Avellini,[c,†,‡] Christophe Lincheneau,[c,†,¶] Serena Silvi,[b,c] 8 Iain A. Wright,[d,¥] Edwin C. Constable[d] and Alberto Credi*[a,b,e] 9 10 11 12 13 Abstract: We describe an operationally advantageous general fluorophores in a variety of applications.[3,4] The use of QDs for 14 methodology to efficiently activate lipoic acid-based compounds - a chemical and biochemical sensing,[ 5 ] medical diagnostics and 15 family of popular surface ligands for semiconductor nanocrystals - by therapy,[6] and as components of photodetectors,[7] light-emitting 16 the use of a borohydride exchange resin, and the use of the activated devices,[8] and solar cells[9] is well established and still actively 17 species to replace the native surface ligands of quantum dots. The investigated. Most fine chemical companies now offer QD 18 procedure enables the phase transfer of the nanocrystals between samples in their catalogues, and flat screen televisions based on 19 polar and aqueous media and, if unsubstituted lipoic acid is used, a QD emission[10] have appeared on the consumer market. 20 facile adjustment of their in a wide range of with Although high quality QDs can be prepared in aqueous 21 varying polarity (from hexane to water). We show that the protocol is media[11] or synthesized in biological/biomimetic systems,[12] the 22 applicable to different types of nanocrystals and a variety of lipoic most reliable and popular synthetic methods exploit reactions of 23 acid-based ligands, and that the resulting quantum dots maintain their inorganic or organometallic precursors in organic solvents in the 24 optical properties - in particular, an intense luminescence - and long presence of appropriate surfactants.[13,14] The QDs afforded by 25 term colloidal stability. such approaches are coated with a layer of highly hydrophobic 26 molecular ligands[15] and are, therefore, (moderately) soluble only 27 in non-polar organic solvents such as toluene, hexane or 28 chloroform. In order to be used in bioimaging and medical therapy, 29 Introduction however, the nanocrystals need to be compatible with and soluble 30 in aqueous media[5,6]. 31 Quantum dots (QDs) are crystals of semiconductor materials In general, the application of these nanomaterials in diverse 32 comprising from some hundreds to a few thousands of atoms, fields of technology implies their facile processing, which 33 with a spherical shape and a diameter typically ranging between is in turn dictated by their solubility in common solvents. Various 34 1 and 15 nm. QDs are endowed with unique size-dependent methods exist to engineer the affinity of semiconductor 35 optical and electronic properties, owing to quantum nanocrystals towards solvents. For example, QDs can be [ 1 , 2 ] 36 confinement. If appropriate semiconducting materials are encapsulated within amphiphilic polymers or peptides, which 37 employed (e.g., CdS, CdSe, CdTe), such quantum dots (QDs) provide a robust coating that preserves the photophysical 38 exhibit a very intense absorption and emission in the UV-visible- properties of the nanoparticles and enhances their colloidal 39 NIR region, and have thus emerged as substitutes for molecular stability.[ 16 ] The increased diameter of the encapsulated 40 nanoparticles, however, may be a problem for biomedical applications.[5,6] Alternatively, the ligands bound at the 41 [a] Dr. M. La Rosa, Prof. Dr. A. Credi [17] 42 Dipartmento di Scienze e Tecnologie Agro-alimentari nanocrystals surface can be altered by chemical reactions. 43 Università di Bologna, Viale Fanin 50, 40127 Bologna, Italy The post-synthetic modification of the surface capping layer also 44 E-mail: [email protected] enables the connection of functional molecular units to the QD, Web: www.credi-group.it leading to the development of hybrid nanomaterials with 45 [b] Dr. M. La Rosa, Dr. S. Silvi, Prof. Dr. A. Credi [4–6,17] 46 CLAN - Center for Light Activated Nanostructures emerging properties. 47 Università di Bologna and Consiglio Nazionale delle Ricerche A more straightforward methodology consists in covering the 48 Via Gobetti 101, 40129 Bologna, Italy surface of the nanocrystals with functional ligands by [c] Dr. T. Avellini, Dr. C. Lincheneau, Dr. S. Silvi replacement of the native ligands. Unfortunately the 49 Dipartimento di Chimica “G. Ciamician”

50 Università di Bologna, Via Selmi 2, 40126 Bologna, Italy photophysical properties as well as the chemical and 51 [d] Dr. I. A. Wright, Prof. Dr. E. C. Constable photochemical stability of the QDs are usually degraded upon 52 Department of Chemistry, University of Basel exchange of the capping ligands.[ 18 , 19 ] Nevertheless, Spitalstrasse 51, 4056 Basel, Switzerland dihydrolipoic acid (DHLA) and related compounds, because of 53 [e] Prof. Dr. A. Credi 54 ISOF-CNR, Via Gobetti 101, 40129 Bologna, Italy the presence of two thiol anchoring groups, are able to form 55 [†] These authors contributed equally to the work. robust monolayers on the surface of inorganic nanocrystals, and 56 [‡] Present address: Istituto Italiano di Tecnologia are frequently used as QD caps.[15,17,20,21,22] Via Morego 30, 16163 Genova, Italy DHLA is obtained from lipoic acid (LA) upon rupture of the 57 [¶] Present address: Université Grenoble Alpes [23] 58 INAC-SyMMES, F-38054 Grenoble Cedex 9, France disulfide bond of its 1,2-dithiolane moiety (Scheme 1). 59 [¥] Present address: Department of Chemistry 60 Loughborough University, Leicestershire LE11 3TU, UK 61 62 63 64 65 FULL PAPER

1 2 3 4 Scheme 1. Chemical reduction of lipoic acid-based compounds to the 5 dihydrolipoic derivatives. 6 7 8 The conversion of LA to DHLA moieties is commonly carried [18-22,24] 9 out in solution using NaBH4 as a reductant. This route, 10 though effective, requires careful preparation, storage, and 11 handling of the DHLA-based ligands under inert atmosphere to

12 prevent reoxidation of the bis-thiol to disulfide. Moreover, NaBH4 13 cannot be used with functional ligands that contain moieties 14 sensitive to Na+ (e.g., receptors for metal ions). A method 15 relying on UV irradiation to cleave the S–S bond of LA-type 16 ligands was reported.[25] Although this route avoids the use of 17 chemical reductants and enables ligand activation and QD 18 functionalization to be performed in a single step, it requires a UV Scheme 2. (a) Reaction between lipoic acid and borohydride exchanged resin 19 reactor and, more importantly, it cannot be employed for capping beads, leading to the chemisorbed dihydrolipoic species (DHLA–). (b) Extraction 20 ligands bearing photosensitive units. of DHLA– from the resin by anion displacement with a salt (M+X–). 21 We recently devised[26] a procedure for efficiently converting 22 lipoic acid into the active DHLA derivative, and subsequently 23 using it to replace the native caps of QDs. Here we report a 24 thorough description of this methodology, which involves the 25 chemical reduction and exchange of LA compounds within a 26 borohydride-loaded resin. In particular, we have tested the 27 procedure on different types of ligands and nanocrystals, and we 28 have investigated the changes in the photophysical properties 29 and colloidal stability of the resulting quantum dots. We have also 30 studied the effect of different reactants used to extract the 31 reduced ligands from the resin on the solubility of the nanocrystals 32 in solvents of varying polarity. 33 34 35 Results and Discussion 36 Figure 1. Absorption spectral changes measured for a 1.610–2 M lipoic acid – solution (a) upon addition of the BER (2 equiv of BH4 with respect to LA) (grey 37 Reduction of the ligand. As discussed above, the dihydrolipoic curves). No more changes were detected after 30 min stirring and the spectrum 38 moiety capable of binding the QD surface can be obtained by (b) was obtained. Treating this solution with NaOH (2 equiv with respect to LA) 39 borohydride-promoted reduction of the corresponding lipoic acid and 30 min stirring afforded the spectrum (c). Conditions: MeOH, room temperature. 40 compound. In our method the use of NaBH4 is avoided and the 41 reduction of the 1,2-dithiolane unit is achieved with a borohydride- 42 loaded ion-exchange resin (BER),[27] as depicted in Scheme 2. 43 BERs are commercially available or can be prepared by mixing After removal of the layer, the beads were washed 44 an anion exchange resin, such as Amberlite® IRA-400, with an with fresh methanol to remove unreacted LA and borohydride [28] products. The DHLA species were extracted from the resin by 45 of NaBH4 and stirring for a few hours. In our 46 case, acid titrations revealed that the resin contained typically 2.5 treatment with a MeOH solution of a metal or salt: the – – – X anions displace DHLA from the resin beads (which are 47 mmol of BH4 per gram of polymer. 48 The conversion of LA to the DHLA anion was performed by successively recovered), and a methanol solution of the active + + – 49 adding the BER beads to a MeOH solution of lipoic acid, with a ligand as its M salt is obtained. We found that the role of M X is – conveniently played by such as NaOH or TMAOH 50 BH4 /LA molar ratio of about 2:1, and stirring for 30 minutes. The 51 process could be conveniently monitored by absorption (TMA = tetramethylammonium). 52 spectroscopy, by following the decrease of the disulfide Indeed, the addition of NaOH to the resin (2 equiv 53 absorption band of lipoic acid at 330 nm (Figure 1).[ 29 ] The with respect to the initial amount of LA) caused an immediate and 54 absorbance in the 220-260 nm region – which is initially out of significant absorption increase between 220 and 250 nm, 55 scale – also decreases. This observation indicates that DHLA, whereas only a minor recovery of the band at 330 nm took place. 56 which exhibits a strong absorption feature in this spectral The signal at 330 nm continued to increase upon stirring for ca. 57 region,[29] is chemisorbed on the cationic resin (Scheme 2a). 20 min (Figure 1, curve c), however reaching a value much 58 59 60 61 62 63 64 65 FULL PAPER

1 smaller than the initial LA band (curve a). These observations 2 indicate that the addition of NaOH causes the quick release of 3 DHLA from the resin; on a longer time scale, a minor amount of 4 unreacted lipoic acid was also released. Similar results were 5 obtained when TMAOH was employed in the place of NaOH. 6 The increase of the absorption intensity at 330 nm was used 7 to estimate the amount of unreacted lipoic acid, whereas the 8 amount of DHLA released in the solution was determined with 9 Ellman’s reagent.[ 30 ] Under our experimental conditions (room 10 temperature; 30 min stirring with BER, addition of 2 equiv of 11 NaOH and 30 min stirring), the DHLA yield in the MeOH solution 12 was 29%; 7% of unreacted LA was also extracted. 13 It is worthy to point out that the purification procedures usually 14 carried out after the reduction of lipoic acid with NaBH4 in [24] 15 homogeneous solution are not needed. In fact the reactant in 16 excess can be separated from the solution simply by decantation [18–22,24] 17 of the resin. Moreover, previous procedures typically 18 involved the preparation of stock quantities of reduced ligand, 19 which would be stored under inert atmosphere at low temperature 20 until used. With such a handy protocol, the amount of DHLA just 21 necessary for a given QD batch can be prepared and readily utilized for the cap exchange. Figure 2. Phase transfer of the hydrophobic QDs in a polar solvent promoted 22 by cap exchange with DHLA ligands. The photographs refer to an experiment Ligand exchange and phase transfer of the QDs. A biphasic 23 in which CdSe-5ZnS core-shell nanocrystals (dcore = 3.6 nm) were employed. 24 liquid system was prepared at room temperature by mixing a The images on the left were taken under ambient light, while those on the right were obtained upon near-UV excitation. 25 hexane solution containing the hydrophobic QDs (capped with 26 long-chain phosphine, phosphine , amine and/or carboxylic 27 derivatives, see Experimental Section) and a methanol solution containing the DHLA ligand. Upon stirring the , an efficient – + 28 Table 1. Spectroscopic properties (H2O, r.t.) of different DHLA Na capped QDs obtained by ligand exchange and phase transfer. Extracting agent: NaOH. 29 ligand exchange occurred and the QDs were transferred to the methanol solution, as indicated by the color change (Figure 2, left 30 [18a,20–22] abs em photographs). The phase transfer can be more clearly d s em,n em,w QD (abs) (em) 31 nm[c] nm[d] [h] [i] 32 appreciated by looking at the QD luminescence upon near-UV nm[e] nm[f] 33 illumination (Figure 2, right photographs). The weak bluish glow 1 CdS[a] 2.8 –– 381 (+9) [g] 34 visible in the hexane layer after the cap exchange arises from the fluorescence of the hydrophobic ligands that remain in the less 35 2 CdS[a] 4.1 –– 420 (+6) [g] 36 polar phase after the exchange. 3 CdSe[a] 2.6 –– 528 (+2) [g] 37 In alternative, the hydrophobic QDs in solid form were added to the DHLA solution; a rapid dispersion of the nanocrystals in 38 4 CdSe[a] 3.8 –– 583 (0) [g] MeOH was observed, witnessing again the efficient replacement 39 of the native hydrophobic ligands with DHLA. 5 CdSe- 2.7 1.6 545 (+3) 562 (+2) 0.14 0.044 40 [b] The discoloured hexane layer was removed if present, and 5ZnS 41 traces of hydrophobic ligands and unreacted nanocrystals were 42 6 CdSe- 3.4 0.9 562 (–2) 585 (–1) 0.34 0.18 washed away by treating the methanol suspension with fresh 3ZnS[b] 43 hexane. After evaporation of MeOH under vacuum, the QDs 44 7 CdSe- 3.6 1.6 572 (0) 597 (–2) 0.23 0.081 capped with DHLA–Na+ were dissolved in water. Large 45 5ZnS[b] aggregates were separated from the aqueous suspension using 46 a syringe filter, and unreacted DHLA was removed by means of 3 8 CdSe- 3.7 1.2 576 (–2) 606 (–1) 0.18 0.06 47 [b] / cycles with a 30 kDa centrifugal filter.[31] 4ZnS 48 Typically, a relatively concentrated (5-10 M) solution of QDs in 49 9 CdSe- 4.1 0.9 587 (–1) 605 (0) water could be eventually obtained. 3ZnS[b] 50

51 [a] Core-only QDs. [b] CdSe-nZnS Core-shell QDs; n denotes the number of 52 ZnS monolayers present in the shell. [c] Core diameter. [d] Shell thickness. [e] 53 abs = Wavelength of the lowest exciton absorption peak in H2O; abs = abs 54 (H2O) – abs (CHCl3). [f] em = Wavelength of the emission band maximum in H2O; em = em (H2O) – em (CHCl3). [g] Not luminescent. [h] Emission quantum 55 yield of QDs with native ligands in CHCl3 or hexane. [i] Emission quantum yield 56 of QDs capped with DHLA–Na+ in water. 57 58 59 60 61 62 63 64 65 FULL PAPER

1 particular, CdSe-ZnS core-shell QDs yielded clear water 2 which were stable for at least 3 months. Transmission electron 3 microscopy (TEM) and dynamic light scattering (DLS) 4 measurements (Figure 3) showed that the cap exchange does not 5 affect the size and morphology of the QDs and no aggregation 6 takes place. Only a slight shift in the absorption and emission 7 peak wavelengths was observed with respect to the starting QDs, 8 indicating that the spectroscopic properties of the final products 9 are preserved (Figures 4 and 5). The luminescence quantum yield 10 of the final nanocrystals in aqueous solution was 30-50% of that [18bd] 11 of the native hydrophobic QDs, in line with literature reports. [18b,32] 12 As observed in earlier studies, CdS and CdSe nanocrystals 13 devoid of a ZnS shell were not emissive after phase transfer. 14 15 16 17 18 19 20 21 22 23 24 25 26 27 Figure 3. Effect of the cap exchange on the size and morphology of the QDs. 28 Top: TEM Micrographs of CdSe-4ZnS QDs (entry 8 in Table 1) TOP/TOPO capped (a) and DHLA–Na+ capped (b) Scale bar, 50 nm. Bottom: Differential 29 (red bars) and cumulative (blue lines) number distribution of particle size, 30 obtained from DLS data, of CdSe QDs (entry 3 in Table 1) TOP/TOPO capped 31 in chloroform (c) and DHLA–Na+ capped in water (d). 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 Figure 4. Absorption (full lines, left scale) and emission (λexc = 420 nm; dashed 49 lines, right scale) spectra of TOP/TOPO CdSe-5ZnS QDs (entry 5 in Table 1) in 50 CHCl3 (black), and of the same QDs in aqueous solution after cap exchange Figure 5. a) Absorption (full line, left scale) and emission (λexc = 485 nm; dashed 51 with DHLA–Na+ (red) or DHLA–TMA+ (blue). Inset: photograph of the DHLA–Na+ line, right scale) spectra of CdSe-3ZnS QDs (entry 6 in Table 1) TOP/TOPO – + 52 capped QDs in water under ambient light (left) and upon UV excitation (365 nm). capped in CHCl3 (black) and DHLA Na capped in H2O (blue). b) Absorption (full line) and emission (λexc = 480 nm; dashed line) spectra of CdSe-5ZnS QDs – + 53 (entry 7 in Table 1) TOP/TOPO capped in CHCl3 (black) and DHLA Na capped 54 in H2O (red). c) Absorption (full line) and emission (λexc = 450 nm; dashed line) – + 55 This method enabled us to produce water soluble DHLA Na spectra of CdSe-3ZnS QDs (entry 9 in Table 1) TOP/TOPO capped in CHCl3 capped nanocrystals of different structure and size (Table 1). In (black) and DHLA-PEG400 capped in H2O (green). Adapted by permission of the 56 Royal Society of Chemistry.[26a] 57 all cases the phase transfer was complete within 1 min. In 58 59 60 61 62 63 64 65 FULL PAPER

1 We found that the reduction with the borohydride-loaded resin displace DHLA– from the resin. and 2 is also effective for ligands consisting of a 1,2-dithiolane moiety afforded an incomplete extraction of the DHLA ligand and 3 attached to a oligo(ethylene glycol) hydrophilic domain,[20,21] such consequently a partial cap exchange occurred. Triflate and 4 as LA-TEG and LA-PEG400. As these compounds do not contain perchlorate salts allowed a complete extraction and consequently 5 ionizable units, the reduced DHLA-based ligand does not stick to afforded cap-exchanged nanocrystals endowed with stability in 6 the resin; hence, the activated ligand can be collected without the water and photophysical properties similar to those prepared 7 extraction step described in Scheme 2b and used in the phase using sodium . Sodium also led to a complete 8 transfer step. Figure 5c shows the absorption and luminescence cap exchange but a longer stirring time was required. 9 spectra of CdSe-ZnS QDs obtained by ligand exchange with DHLA-PEG400. 10 Table 2. Photophysical properties (H2O, room temperature) of CdSe-5ZnS 11 (entry 7 in Table 1) capped with DHLA–Na+ extracted from the BER using 12 different sodium salts. 13 + – Phase Solubility abs em 14 The stability of DHLA capped QDs prepared with the above Na X [a] [b] [c] transfer in H2O nm nm 15 described route on the long term was evaluated. For example, a 16 dilute solution (130 nM) of the nanocrystals described in Figure NaCl No No –– –– 17 5b in deionized water was stored in a refrigerator at 5° C, and the NaBH4 No No –– –– 18 absorption and luminescence spectra were monitored over 3 19 weeks (Figure 6). No precipitation was observed, although the NaBr Complete Yes –1 –2

20 emission quantum yield decreased from 0.081 to 0.05 during the Na2CO3 Partial Yes –1 –2 21 first two weeks, in line with literature reports for DHLA-capped NaCH3COO Partial Yes –2 –2 22 QDs.[33] 23 NaCF3SO3 Complete Yes –2 –1

24 NaClO4 Complete Yes 0 –1

25 NaOH Complete Yes 0 –2 26 27 [a] The QDs are considered soluble if they form a homogeneous solution at the 28 0.5-1.0 M concentration level. [b] abs = abs (H2O) – abs (CHCl3); abs = 572 29 nm; wavelength of the lowest exciton absorption peak in H2O; [c] em = 597 nm; 30 wavelength of the emission band maximum in H2O; em = em (H2O) – em (CHCl3). 31 32 Successively, we investigated the effect of the M+ cation by 33 exchanging the ligand on CdSe-ZnS QDs using different alkali 34 metal or ammonium hydroxides in the extraction step (Scheme 35 2b). The results are summarized in Table 3. In all cases the phase 36 transfer was successful, the resulting QDs were soluble in water 37 and their spectroscopic properties were preserved. 38

39 Figure 6. Absorption (full lines, left scale) and emission (exc = 480 nm; dashed Table 3. Photophysical properties (H2O, room temperature) of CdSe-5ZnS 40 lines, right scale) spectra of CdSe-5ZnS QDs (entry 7 in Table 1, 130 nM) (entry 7 in Table 1) capped with DHLA–Na+ extracted from the BER using 41 capped with DHLA–Na+ in water, freshly prepared (black) and after 21 days of different metal or ammonium hydroxides. 42 storage at 5 °C (red). The inset shows the evolution of the luminescence quantum yield over time. Adapted by permission of the Royal Society of Phase Solubility abs em [26a] + – 43 Chemistry. M OH [a] [b] [c] transfer in H2O nm nm 44 45 LiOH Complete Yes 0 –1 46 Effect of the reactant employed to collect DHLA from the BER. NaOH Complete Yes 0 –2 47 To gain more insight about the role of the M+X– species (Scheme 48 2b) in determining the properties of the final phase-transferred KOH Complete Yes –4 –3 49 QDs, we carried out several cap exchange reactions employing TMAOH[d] Complete Yes 0 0 50 different metal or ammonium salts or hydroxides to extract DHLA TBAOH[e] Complete Yes 0 –5 51 from the resin. 52 In a first instance, we explored the influence of the nature of [a] The QDs are considered soluble if they form a homogeneous solution at the – 53 the X anion by using different sodium salts in the extraction step 0.5-1.0 M concentration level. [b] abs = abs (H2O) – abs (CHCl3); abs = 572 54 (Scheme 2b) and successively performing phase transfer on nm; wavelength of the lowest exciton absorption peak in H2O; [c] em = 597 nm; 55 CdSe-ZnS core-shell nanocrystals. The results are summarized wavelength of the emission band maximum in H2O; em = em (H2O) – em (CHCl3); [d] TMA = tetramethylammonium; [e] TBA = tetra(n-butyl)ammonium. 56 in Table 2. No phase transfer was observed with NaCl or NaBH4, 57 suggesting that and borohydride anions are unable to 58 59 60 61 62 63 64 65 FULL PAPER

1 Extraction with salts of transition metal (CuII triflate) and TBA = tetra(n-butyl)ammonium; TOA = tetra(n-octyl)ammonium; CTA = lanthanide (LaIII triflate) ions was also attempted; phase transfer cetyltrimethylammonium. [b] Relative dielectric constant. [c] Hydroxide, , 2 perchlorate or triflate salts. [d] bromide salt. [e] Native QDs TOP/TOPO capped. 3 in methanol was observed but the QDs could not be dissolved in 4 water. As the hydration of Cu2+ and Ln3+ ions is strongly In all the dispersions listed in Table 4 the nanocrystals 5 exothermic, it can be hypothesized that the lack of solubility of the maintained their spectroscopic properties; the shifts of the 6 QDs in water is due to a very strong binding of and, in absorption and emission peaks with respect to the native QDs did 7 particular, lanthanum ions to the carboxylate residues on the QD not exceed 5 nm (see, e.g., Figure 7). 8 surface. Moreover, the QDs suspended in methanol exhibited no 9 luminescence, most likely because these ions can enable [34] 10 electron-transfer quenching pathways. The same result was II 11 observed when [Ru (bpy)3](ClO4)2 (bpy = 2,2'-bipyridine) was [35] 12 used in the extraction step. 13 Modulation of the QDs solubility. During the experiments 14 performed to investigate the influence of the extraction step 15 (Scheme 2b) in the ligand exchange process (Figure 2), we + – 16 realized that the nature of the M X species can significantly affect 17 the solubility pattern of the final QDs in different solvents. We thus monitored the solubility of different phase-transferred QDs in a Figure 7. Photographs of 0.5 M CdSe-5ZnS QDs (entry 7 in Table1) capped 18 – + range of solvents of varying polarity, from hexane to water. with DHLA TBA in methanol (a), acetonitrile (b), DMSO (c) and water (d) under 19 ambient light (top) and UV light (365 nm, bottom). Adapted by permission of the The results of these tests, summarized in Table 4, show that 20 Royal Society of Chemistry.[26a] 21 using the same cap exchange procedure and the same capping 22 agent (lipoic acid) it is possible to tune the solubility of the QDs in 23 different solvents simply by changing the type of salt used in the A rationalization of the solubility pattern shown in Table 4 can 24 resin extraction step. DHLA-capped QDs possess a negatively be attempted considering that dissolution thermodynamics is + 25 charged surface wherein carboxylate residues are paired with M determined by the combination of endothermic (breaking of the 26 counterions; evidently, the nature of such cations affects the interactions between QDs in the solid phase and between solvent 27 ability of the QDs to be dispersed in different solvents. Although molecules) and exothermic ( of the polyanionic QDs and counterion effects in nanocrystals capped with ionic ligands have their counter-cations) processes. The large solvation enthalpies 28 [36] 29 been described, reports on the role of counterions for tuning the of alkali cations and quaternary ammonium ions with short alkyl 30 solubility of charged QDs in polar solvents are unprecedented. groups (e.g. TMA+) in polar solvents may explain why the 31 corresponding DHLA–M+ QDs are soluble in methanol and/or 32 Table 4. Solubility chart in various solvents at room temperature of CdSe-3ZnS water. Although it may be anticipated that tetraalkylammonium + [a] 33 QDs (entry 9 in Table 1) capped with DHLA and different M countercations. ions with longer chains could render the QDs compatible with less polar organic solvents, a clear trend was not observed. Specific 34 Solvent ([b]) 35 effects may be significant in particular cases: for example, cation-

 interactions of the ions with the aromatic solvent

36 molecules could explain the solubility of DHLA–Li+ QDs in toluene. 37 M+ 38 It is worth to note that QDs capped with DHLA– and TBA+ ions are 39 soluble in a large variety of solvents, ranging from chloroform to

THF (7.58)

Water (80.2)

DMSO (46.5)

Hexane (1.89)

Toluene (2.38) Acetone (20.7) water (Table 4, Figure 7). This finding indicates that such bulky 40 Methanol (32.7)

Chloroform (4.81) Acetonitrile (35.9) 41 cations afford a good balance, in terms of solid phase and 42 Li+ [c]    solvation energies, in solvents as different as water and 43 Na+ [c]    chloroform. 44 K+ [c]   45 Mg2+ [c] 46 Ca2+ [c] Conclusions

47 + [c] NH4  48 In this work we have described a general methodology for the + [c] 49 TMA    chemical activation of lipoic acid-type ligands and their 50 TEA+ [c]    successive utilization for phase transfer of semiconductor 51 TBA+ [c]        quantum dots in polar and aqueous solvents. Our results show 52 TOA+ [c]  that the protocol is applicable to different types of nanocrystals 53 CTA+ [d]     and a variety of dithiolane-based ligands, and that the resulting QDs maintain their optical properties (in particular, an intense 54 nQDs[e]     55 luminescence) and long-term colloidal stability. 56 [a] Circles indicate that the QDs form a homogeneous solution at the 0.5-1.0 M The procedure is handy and has several practical 57 concentration level. TMA = tetramethylammonium; TEA = tetraethylammonium; advantages: 58 59 60 61 62 63 64 65 FULL PAPER

1 (i) the resin-supported reactant can be separated from the Formvar resin film supported on conventional copper microgrids 2 reaction product simply by decantation or filtration; and the sample was dried under vacuum. DLS measurements 3 (ii) the reduction can be performed at room temperature and in were performed with a Malvern Nano ZS instrument equipped 4 aerated conditions; with a 633 nm laser diode. The measurements were carried out 5 (iii) the reaction time is much shorter (ca. 1 h) than that required at room temperature in air-equilibrated solutions placed in 1 cm 6 when NaBH4 is used (ca. 5 h); quartz cuvettes. Electronic absorption spectra were measured at 7 (iv) the ligand can be quickly reduced, just in the amount room temperature on air-equilibrated solutions of the samples 8 necessary for the successive cap exchange, thus avoiding the contained in quartz cuvettes (1-cm optical path length). The 9 preparation of DHLA stock solutions and their storage at low concentration of the solutions was typically 10–7 M. Absorption 10 temperature under inert atmosphere. spectra in the 190-1100 nm range were recorded with a Perkin 11 Moreover, in the case of ligand exchange with lipoic acid, our Elmer 45 spectrophotometer. The precision on the wavelength 12 method enables the precise modulation of the QD solubility in a values was ± 1 nm. Luminescence spectra in the 250-900 nm 13 wide variety of organic solvents through the choice of the range were recorded with a Perkin Elmer LS50 or an Edinburgh 14 countercations associated with the carboxylate moieties of the Instruments FLS920 spectrofluorimeter equipped with a 15 surface-attached DHLA species. The fact that QDs capped with Hamamatsu R928 photomultiplier. Luminescence spectra were 16 the same organic ligand exhibit significantly different solubility recorded at room temperature (ca. 295 K) on solutions of samples 17 properties is indeed a remarkable finding. contained in quartz cuvettes (1-cm optical path length). 18 Engineering the surface capping layer of QDs constitutes a Luminescence quantum yields were determined with the optically 19 versatile approach to modulate their properties or to implement dilute method using Rhodamine 6G ( = 0.94 in EtOH) as a [37] 20 new functions. Thus, the development of simple and efficient standard. 21 routes for the chemical modification of the nanocrystals surface is Synthesis of the Quantum Dots. Hydrophobic core CdS and 22 desirable. In particular, the ability to predetermine the solubility of CdSe semiconductor nanocrystal quantum dots were prepared [14cd] 23 QDs in common solvents without affecting their photophysical following the procedures published by Peng and coworkers 24 behaviour is a necessary step forward toward the application of with minor modifications. Core-shell CdSe-ZnS QDs were 25 these nanomaterials. prepared by overcoating a CdSe core with a ZnS shell using either [14e] [14b,38] 26 SILAR or one-time-precursors-injection approaches. The 27 core diameter and the QD concentration were estimated Experimental Section according to published methods.[ 39 ] The shell thickness was 28 [14be,38] 29 estimated as reported in each synthetic protocol. The 30 Reagents and Solvents. Lipoic acid (LA, (±)-α-Lipoic acid, resulting nanocrystals were covered with TOPO (tris-[n- octylphosphine]oxide), TOP (tris-[n-octyl]phosphine), OA (oleic 31 ≥98%), sodium borohydride (NaBH4, ≥98%), anion exchange 32 resin Cl– loaded (Amberlite® IRA-400 chloride form), cadmium acid), ODA (n-octadecylamine) and/or HDA (n-hexadecylamine) 33 oxide (99.99%), selenium powder (99.5%, 100 mesh), sulphur as passivating surface ligands to prevent particle aggregation. 34 powder (99.98%), tris(n-octyl)phosphine oxide (TOPO, 90%), Preparation of the Borohydride-Exchanged Resin. The tris(n-octyl)phosphine (TOP, 97%), n-hexadecylamine (HDA, synthesis of the BH4-resin was carried out according to a 35 [28] 36 98%), n-octadecylamine (ODA, 97%), oleic acid (OA, ≥99%), published procedure. Briefly, 4 g of Amberlite® IRA-400 37 diethyl (1 M solution in n-heptane), (NaOH, (chloride form) resin were placed in a 100 mL one-neck round- bottom flask equipped with a stirring bar. A water solution of 38 ≥98%), monohydrate (LiOH, >99.0%), NaBH (920 mg in 40 mL) was added and the neck was sealed 39 hydroxide (KOH, ≥98%), tetramethylammonium 4 with a rubber seal. The mixture was gently stirred for at least 3 40 hydroxide (TMAOH, ~97%), tetraethylammonium perchlorate hours in order to allow the complete exchange of the chloride 41 (TEAClO4, 99%), tetraethylammonium nitrate (TEANO3, >99%), anions with borohydride. The resin was filtered and washed with 42 tetra(n-octyl)ammonium hydroxide (TOAOH, 20% in methanol), water to remove the excess of sodium borohydride and sodium 43 cetyltrimethylammonium bromide (CTABr, ≥99.0%), and 5,5'- chloride released during the exchange reaction. The resin beads 44 dithiol-bis(2-nitrobenzoic acid) (DTNB, ≥98%) were purchased were then dried under reduced pressure. The amount of BH – per 45 from Sigma Aldrich and Fluka, and were used without further 4 g of resin was estimated via acid titration. 46 purification. Tetra(n-butyl)ammonium hydroxide (TBAOH, 25% in Ligand Reduction. In a 4-mL glass vial a solution of lipoic acid 47 methanol) was purchased from J. T. Baker Chemical Co. (2.6610–5 mol in 500 L of MeOH) was mixed with 19 mg of BH – 48 Synthetic grade hexane, toluene, CHCl3, acetone and methanol 4 resin (2.5 mmol BH – per g). The mixture was stirred at 400 rpm 49 were purchased from Sigma-Aldrich. Spectroscopic grade 4 for at least 30 min. The methanol layer was removed and the 50 tetrahydrofuran, acetonitrile and dimethyl sulfoxide were beads were washed 3 times with 500 L of methanol to remove 51 purchased from Merk-Uvasol. A Milli-Q Millipore system was used unreacted lipoic acid and hydrolyzed borohydride products. 52 for the purification of water (resistivity ≥ 18 M·cm). Millipore Methanol (500 L) and M+X– (from 1.2 to 4 eq with respect to the 53 Amicon Ultra-0.5 mL centrifugal filters (30000 Da cut off) were BH – content) were then added to the beads. The mixture was 54 purchased from Sigma-Aldrich. 4 Instrumental measurements. TEM experiments were carried stirred for at least 30 min in order to extract the reduced lipoic acid 55 + – out with a Philips CM 100 transmission electron microscope from the resin. In the case of hydroxide salts, 1.2 eq of M OH 56 and 30 min of stirring were enough to extract the reduced ligand 57 operating at 80 kV. A drop of QD solution was deposited on a 58 59 60 61 62 63 64 65 FULL PAPER

1 from the resin; perchlorate, nitrate and bromide salts required a obtaining a suspension. All the successive isolation and 2 larger amount (up to 4 eq) and a longer stirring time (up to 3 hours). purification steps were performed as described above. 3 The beads were washed two times with 200 L of fresh methanol Determination of the DHLA yield. The amount of DHLA in 4 to recover as much reduced ligand as possible. The methanol solution after extraction from the BER was determined by 5 fractions were combined and the residual solid was discarded. In measuring the concentration of its thiol groups using Ellman’s 6 the case of ligands consisting of a lipoic acid moiety attached to a reagent (5,5'-dithiol-bis(2-nitrobenzoic acid), DTNB).[30] 10 L of a [18] – + 7 hydrophilic poly(ethylene glycol) chain (LA-TEG, LA-PEG400), MeOH solution containing a DHLA M /LA mixture after extraction 8 the reaction was slower than for parent LA, and no base was from the resin (initial LA concentration, 1.610–2 M) were mixed 9 required to extract the reduced ligand from the resin. with 2 mL of Ellman’s reagent solution (2.110–4 M in 5 mM [29] 10 QD Cap exchange and phase transfer. The desired amount of buffer at pH 7.5). The formation of the reaction 2– 11 QDs (from 1/20000 to 1/30000 QD/lipoic ligand ratio, depending product, the 5-thio-2-nitrobenzoate (TNB ), was monitored 12 on QD size) was dissolved in 1 mL of hexane. The solution was recording the absorbance at 412 nm, and the concentration of the 13 added to the vial containing the reduced ligand (see above) in SH groups was determined using a molar absorption coefficient –1 –1 2– [ 40 ] 14 methanol, thus forming a biphasic mixture. The transfer of QDs of 14150 M cm for TNB at 412 nm. The DHLA 15 from the hexane to the methanol layer was immediately observed. concentration was calculated as [SH]/2. 16 To ensure a complete cap exchange, the biphasic system was 17 stirred for 2-3 hours or, in the case of larger QDs, overnight. The 18 methanol layer appeared clear or turbid, depending on the Acknowledgements 19 counter-cation employed in the previous step (see Table 2). The 20 hexane layer (turned colorless) was removed and the methanol Financial support from the EU (FP7-NMP project Hysens, No. 21 suspension was washed five times with hexane (2 mL) in order to 263091), the Italian Ministry for Education, University and 22 remove unreacted nanocrystals and native hydrophobic ligands. Research (PRIN 2010CX2TLM InfoChem), and the University of The methanol was evaporated under reduced pressure and 23 Bologna is gratefully acknowledged. We also thank the Università resulting dried QDs were dissolved in proper solvents for further 24 Italo-Francese for a Vinci fellowship to M. L. R. 25 studies. In the case of water solutions, the mixture was first passed through a syringe filter (0.46 m pore size) to remove 26 Keywords: CdSe • nanoparticle • phase transfer • possible large aggregates and was successively purified with 3 27 semiconductor nanocrystal • solubility 28 cycles of dilution/concentration with a centrifugal filter (Amicon 29 Ultra-0.5 mL, 30 kDa, 7000 rpm, 12 minutes) to eliminate the

30 excess of free ligand. Alternatively, hydrophobic QDs were added as solid/paste to the solution of reduced ligand in methanol, thus 31 32 R. Manders, J. Xue, P. H. Holloway, L. Qian, Nat. Photon. 2015, 9, 259- 266. 33 [1] Semiconductor Nanocrystal Quantum Dots, A. L. Rogach, (Ed.), [9] a) P. V. Kamat, J. Phys. Chem. C 2008, 112, 18737-18753; b) S. Rühle, 34 Springer-Verlag, Wien, 2008. M. Shalom, A. Zaban, ChemPhysChem 2010, 11, 2290-2304; c) G. H. 35 [2] Top. Curr. Chem. 2016, 347, 1-174 (Special issue on Photoactive Carey, A. L. Abdelhady, Z. Ning, S. M. Thon, O. M. Bakr, E. H. Sargent, 36 Semiconductor Nanocrystal Quantum Dots - Fundamentals and Chem. Rev. 2015, 115, 12732-12763. 37 Applications). [10] X. Dai, Y. Deng, X. Peng, Y. Jin, Adv. Mater., in press, DOI: 38 [3] Quantum Dot Sensors: Technology and Commercial Applications, J. F. 10.1002/adma.201607022 39 Callan, F. M. Raymo (Eds.), Pan Stanford Publishing, Singapore, 2012. [11] a) V. Lesnyak, N. Gaponik, A. Eychmüller, Chem. Soc. Rev. 2013, 42, 40 [4] a) U. Resch-Genger, M. Grabolle, S. Cavaliere-Jaricot, R. Nitschke, T. 2905-2929; b) M. Ulusoy, R. Jonczyk, J. G. Walter, S. Springer, A. Nann, Nat. Methods 2008, 5, 763-775; b) B. Hötzer, I. L. Medintz, N. Lavrentieva, F. Stahl, M. Green, T. Scheper, Bioconjugate Chem. 2016, 41 Hildebrandt, Small 2012, 8, 2297-2326. 27, 414-426. 42 [5] a) I. L. Medintz, H. T. Uyeda, E. R. Goldman, H. Mattoussi, Nat. Mater. [12] J. Zhou, Y. Yang, C. Zhang, Chem. Rev. 2015, 115, 11669-11717. 43 2005, 4, 435-447; b) R. Gill, M. Zayats, I. Willner, Angew. Chem. Int. Ed. [13] a) Y. Yin, A. P. Alivisatos, Nature 2005, 437, 664-670; b) C. de Mello 44 2008, 47, 7602-7625; c) T. Delgado-Perez, L. M. Bouchet, M. de la Donegá, P. Liljeroth, D. Vanmaekelbergh, Small 2005, 1, 1152-1162; c) 45 Guardia, R. E. Galian, J. Pérez-Prieto, Chem. Eur. J. 2013, 19, 11068- J. Park, J. Joo, S. G. Kwon, J. Jang, T. Hyeon, Angew. Chem. Int. Ed. 46 11076; d) S. Silvi, A. Credi, Chem. Soc. Rev. 2015, 44, 4275-4289; e) K. 2007, 46, 4630-4660; d) P. Reiss, M. Carrière, C. Lincheneau, L. Vaure, 47 D. Wegner, N. Hildebrandt, Chem. Soc. Rev. 2015, 44, 4792-4834; f) N. S. Tamang, Chem. Rev. 2016, 116, 10731-10819. Hildebrandt, C. M. Spillmann, W. R. Algar, T. Pons, M. H. Stewart, E. Oh, [14] a) C. B. Murray, D. J. Norris, M. G. Bawendi, J. Am. Chem. Soc. 1993, 48 K. Susumu, S. A. Diaz, J. B. Delehanty, I. L. Medintz, Chem. Rev. 2017, 115, 8706-8715; b) M. A. Hines, P. Guyot-Sionnest, J. Phys. Chem. 1996, 49 117, 536-711. 100, 468-471; c) Z. A. Peng, X. Peng, J. Am. Chem. Soc. 2001, 123, 183- 50 [6] a) P. Zrazhevskiy, M. Sena, X. Gao, Chem. Soc. Rev. 2010, 39, 4326- 184; d) W. W. Yu, X. Peng, Angew. Chem. Int. Ed. 2002, 41, 2368-2371; 51 4354; b) T. L. Doane, C. Burda, Chem. Soc. Rev. 2012, 41, 2885-2911; e) J. J. Li, Y. A. Wang, W. Guo, J. C. Keay, T. D. Mishima, M. B. Johnson, 52 c) J. D. Meyers, T. Doane, C. Burda, J. P. Basilion, Nanomedicine 2013, X. Peng, J. Am. Chem. Soc. 2003, 125, 12567-12575. 53 8, 123-143; d) A. F.-J. Jou, C.-H. Lu, Y.-C. Ou, S.-S. Wang, S. L. Hsu, I. [15] a) M. Green, J. Mater. Chem. 2010, 20, 5797-5809; b) M. A. Boles, D. 54 Willner, J. A. A. Ho, Chem. Sci. 2015, 6, 659-665. Ling, T. Hyeon, D. V. Talapin, Nat. Mater. 2016, 15, 141-153. 55 [7] F. P. G. de Arquer, A. Armin, P. Meredith, E. H. Sargent, Nat. Rev. Mater. [16] See, e.g.: J. Ostermann, J.-P. Merkl, S. Flessau, C. Wolter, A. Kornowksi, 2017, 2, 16100. C. Schmidtke, A. Pietsch, H. Kloust, A. Feld, H. Weller, ACS Nano 2013, 56 [8] a) Y. Shirasaki, G. J. Supran, M. G. Bawendi, V. Bulovic, Nat. Photon. 7, 9156-9167. 57 2013, 7, 13-23; b) Y. Yang, Y. Zheng, W. Cao, A. Titov, J. Hyvonen, J. 58 59 60 61 62 63 64 65 FULL PAPER

1 [26] a) T. Avellini, C. Lincheneau, M. La Rosa, A. Pertegas, H. J. Bolink, I. A. Wright, E. C. Constable, S. Silvi, A. Credi, Chem. Commun. 2014, 50, 2 [17] See, e.g.: a) S. Impellizzeri, B. McCaughan, J. F. Callan, F. M. Raymo, 11020-11022; b) International patent no. WO2014181245. J. Am. Chem. Soc. 2012, 134, 2276-2283; b) M. Oszajca, C. Lincheneau, 3 [27] A. Kirschning, J. Prakt. Chem. 2000, 342, 508-511. M. Amelia, C. Schäfer, K. Szaciłowski, A. Credi, Eur. J. Inorg. Chem. 4 [28] N. M. Yoon, H. J. Lee, J. H. Ahn, J. Choi, J. Org. Chem. 1994, 59, 4687- 2013, 3550-3556; c) W. Wang, A. Kapur, X. Ji, B. Zeng, D. Mishra, H. 5 4688. Mattoussi, Bioconjugate Chem. 2016, 27, 2024-2036. [29] G. Bucher, C. Lu, W. Sander, ChemPhysChem 2005, 6, 2607-2618, and 6 [18] a) J. Aldana, Y. A. Wang, X. Peng, J. Am. Chem. Soc. 2001, 123, 8844- references therein. 7 8850; b) J. A. Kloepfer, S. E. Bradforth, J. L. Nadeau, J. Phys. Chem. B [30] G. L. Ellman, Arch. Biochem. Biophys. 1959, 82, 70-77. 8 2005, 109, 9996-10003; c) A. R. Clapp, E. R. Goldman, H. Mattoussi, [31] B. C. Mei, K. Susumu, I. L. Medintz, H. Mattoussi, Nat. Protoc. 2009, 4, 9 Nat. Protoc. 2006, 1, 1258-1266; d) K. Susumu, E. Oh, J. B. Delehanty, 412-423. 10 J. B. Blanco-Canosa, B. J. Johnson, V. Jain, W. J. Hervey IV, W. R. Algar, [32] a) R. Xie, U. Kolb, J. Li, T. Basché, A. Mews, J. Am. Chem. Soc. 2005, K. Boeneman, P. E. Dawson, I. L. Medintz, J. Am. Chem. Soc. 2011, 133, 11 127, 7480-7488; b) C. Bullen, P. Mulvaney, Langmuir 2006, 22, 3007- 9480-9496; e) S. Tamang, G. Beaune, I. Texier, P. Reiss, ACS Nano 12 3013; c) R. Calzada, C. M. Thompson, D. E. Westmoreland, K. Edme, E. 2011, 5, 9392-9402. 13 A. Weiss, Chem. Mater. 2016, 28, 6716-6723. [19] For an exception, see: J. Aguilera-Sigalat, S. Rocton, J. F. Sánchez- [33] D. Liu, P. T. Snee, ACS Nano 2011, 5, 546-550. 14 Royo, R. E. Galian, J. Pérez-Prieto, RSC Advances 2012, 2, 1632-1638. [34] S. Silvi, M. Baroncini, M. La Rosa, A. Credi, Top. Curr. Chem. 2016, 374, 15 [20] a) H. T. Uyeda, I. L. Medintz, J. K. Jaiswal, S. M. Simon, H. Mattoussi, J. 65. 16 Am. Chem. Soc. 2005, 127, 3870-3878; b) G. Palui, H. B. Na, H. [35] M. Amelia, M. Font, A. Credi, Dalton Trans. 2011, 40, 12083-12088. 17 Mattoussi, Langmuir 2012, 28, 2761-2772. [36] M. V. Kovalenko, M. I. Bodnarchuk, D. V. Talapin, J. Am. Chem. Soc. [21] W. Liu, M. Howarth, A. B. Greytak, Y. Zheng, D. G. Nocera, A. Y. Ting, 18 2010, 132, 15124-15126; b) Q. Feng, L. Dong, J. Huang, Q. Li, Y. Fan, M. G. Bawendi, J. Am. Chem. Soc. 2008, 130, 1274-1284. 19 J. Xiong, C. Xiong, Angew. Chem. Int. Ed. 2010, 49, 9943-9946; c) A. [22] a) I. Yildiz, B. McCaughan, S. F. Cruickshank, J. F. Callan, F. M. Raymo, 20 Nag, D. S. Chung, D. S. Dolzhnikov, N. M. Dimitrijevic, S. Chattopadhyay, Langmuir 2009, 25, 7090-7096; b) I. Yildiz, E. Deniz, B. McCaughan, S. T. Shibata, D. V. Talapin, J. Am. Chem. Soc. 2012, 134, 13604-13615; 21 F. Cruickshank, J. F. Callan, F. M. Raymo, Langmuir 2010, 26, 11503- d) D. M. Balazs, D. N. Dirin, H.-H. Fang, L. Protesescu, G. H. ten Brink, 22 11511. B. J. Kooi, M. V. Kovalenko, M. A. Loi, ACS Nano 2015, 9, 11951-11959. 23 [23] I. C. Gunsalus, L. S. Barton, W. Gruber, J. Am. Chem. Soc. 1956, 78, [37] M. Montalti, A. Credi, L. Prodi, M. T. Gandolfi, Handbook of 24 1763-1766. Photochemistry, 3rd Ed., CRC-Taylor and Francis, Boca Raton, 2006. 25 [24] H. Mattoussi, J. M. Mauro, E. R. Goldman, G. P. Anderson, V. C. Sundar, [38] B. O. Dabboussi, J. Rodriguez-Viejo, F. V. Mikulec, J. R. Heine, H. F. V. Mikulec, M. G. Bawendi, J. Am. Chem. Soc. 2000, 122, 12142- 26 Mattoussi, R. Ober, K. F. Jensen, M. G. Bawendi, J. Phys. Chem. B 1997, 12150. 27 101, 9463-9475. [25] a) G. Palui, T. Avellini, N. Zhan, F. Pan, D. Gray, I. Alabugin, H. Mattoussi, 28 [39] W. W. Yu, L. Qu, W. Gou, X. Peng, Chem. Mater. 2003, 15, 2854-2860. J. Am. Chem. Soc. 2012, 134, 16370-16378; b) N. Zhan, G. Palui, H. [40] P. W. Riddles, R. L. Blakeley, B. Zerner, Anal. Biochem. 1979, 94, 75- 29 Mattoussi, Nat. Protoc. 2015, 10, 859-874. 30 81. 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 FULL PAPER 1 2 3 Entry for the Table of Contents 4 5 6 Layout 2: 7 8 9 FULL PAPER 10 11 Photoactive semiconductor 12 nanocrystals 13 14 Marcello La Rosa, Tommaso Avellini, 15 Christophe Lincheneau, Serena Silvi, 16 Iain A. Wright, Edwin C. Constable and 17 Alberto Credi* 18 The exchange with ligands comprising the 1,2-dithiolane moiety, activated by means Page No. – Page No. 19 of a borohydride-loaded resin, enables the transfer of hydrophobic quantum dots in 20 polar solvents while preserving their spectroscopic properties. If lipoic acid is used An efficient method for the surface 21 as the capping ligand, the solubility of the dots in solvents from hexane to water can functionalization of luminescent 22 be finely tuned by the choice of the countercations associated with carboxylate quantum dots with lipoic acid-based 23 residues present on the nanocrystal surface. ligands 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65